Closed daroczig closed 8 years ago
@RomanTsegelskyi, will you have some time to look into this?
@daroczig, sorry, I was travelling during the holidays, now back in Seattle and will look into all the issues this week
Sounds good, thanks for looking into this.
After some exploration it seems to me that test failure was introduced by the regression in rms
, for which I logged the issue (https://github.com/harrelfe/Hmisc/issues/38). The code we were using was provided as part of #207, so I changed the test example to use official exaples from rms documentation in #240, which might be more robust aganist changes in future.
Thanks @RomanTsegelskyi. Travis reported the rmarkdown
package to be missing and thus the tests failing. Are you looking into this or should I?
It seems that new version of knitr
is defaulting to rmarkdown
. I added it to the list of recommended packages
Thanks a lot!
@RomanTsegelskyi, as per the referenced Hmisc
ticket, it seems to me that the reported issue should not happen -- no? I mean Travis is installing the most recent version of Hmisc
and rms
, but in the ticket it was said that the most recent CRAN version of these packages should not fail. Now I am not sure if that's the case, so I'll leave this to you if any further follow-ups are needed or not.
I still feel that it was connected with incorrect version of the Hmisc
, because I can't reproduce it anymore after updaing Hmisc
Sounds good, thanks for the feedback.
rms
support was added in #207, but a recent update in the package broke functionality. @RomanTsegelskyi, can you please have a look and fix it? See eg https://travis-ci.org/Rapporter/pander/builds/99020762I think an elegant solution might be to check the version of
rms
and support old and new versions (as forggplot2
in https://github.com/Rapporter/pander/commit/ce75329da31e9ea6a784e2fde254815cbb591095), but I'm fine with dropping support for older builds if we can specify min version number for this suggested package.